Cell, Vol. 105, 197–207, April 20, 2001, Copyright 2001 by Cell Press

Apical Localization of wingless TranscriptsIs Required for Wingless Signaling

isms and cell types (reviewed in Bashirullah et al., 1998).These transcripts also encode a variety of proteins,ranging from transcription factors to components of the

Andrew J. Simmonds,* Gilbert dosSantos,*†

Izhar Livne-Bar,§ and Henry M. Krause*‡

*Banting and Best Department of Medical Researchcytoskeleton. Although relatively few localized tran-University of Torontoscripts have been tested for functional relevance, local-Room 312ization has been shown to be important for function inCharles H. Best Institutethe majority of those cases tested. Some well-character-112 College Streetized examples include Actin localization at the leadingToronto, Ontarioedge of migrating fibroblasts (Lawrence and Singer,Canada1986; Kislauskis et al., 1997), oskar localization at the†Program in Developmental Biologyposterior pole of Drosophila oocytes (Ephrussi and Leh-University of Torontomann, 1992), Prospero localization in dividing Drosoph-Canadaila neuroblasts (Broadus et al., 1998), and Ash1 localiza-tion in dividing yeast cells (Long et al., 1997; Takizawaet al., 1997).Summary

The transcript under investigation in this study en-codes a secreted signaling molecule. To date, thereMany developing and adult tissues are comprised ofare relatively few examples of localized transcripts thatpolarized epithelia. Proteins that are asymmetricallyencode signaling molecules and these occur predomi-distributed in these cells are thought to be localizednantly in oocytes. One example with well-documentedby protein trafficking. Here we show that the distribu-functional relevance is the localization of gurken (grk)tion and function of the signaling protein Wingless istranscripts. grk encodes a TGFb protein, and localizedpredetermined by the subcellular localization of itstranslation within the Drosophila oocyte results in di-mRNA. High-resolution in situ hybridization revealsrected secretion to nearby follicle cells (Neuman-Sil-apical transcript localization in the majority of tissuesberberg and Schupbach, 1993). Although there areexamined. This localization is mediated by two inde-examples of signal and receptor molecules that are en-pendently acting elements in the 39 UTR. Replacementcoded by localized transcripts in epithelial cells (sev-of these elements with non- or basolaterally localizingenless for example; Tomlinson et al., 1987), it has yet toelements yields proteins with altered intracellular andbe determined whether or not these transcript localiza-extracellular distributions and reduced signaling activ-tion events have functional significance.ities. This novel aspect of the wingless signaling path-

The subject of this study, wingless (wg), is the proto-way is conserved and may prove to be a mechanismtypical member of the highly conserved Wnt gene family.used commonly for establishing epithelial cell polarity.Wnt genes encode secreted glycoproteins that serve asmajor signaling molecules in a large number of embry-

Introduction onic patterning processes (Wodarz and Nusse, 1998).Wingless protein (WG), and its vertebrate ortholog

The majority of cells, including those of most single- WNT1, function interchangeably in a variety of assayscell organisms, are at least transiently asymmetrical in including the oncogenic transformation of mammaryshape, composition, and structure. Epithelial cells, for cells (Bocchinfuso et al., 1999). This interchangeabilityexample, have well-defined apical and basal mem- is due largely to the striking conservation of downstreambranes specialized for the import and export of different components in the signaling pathway (reviewed in Wo-molecules. One of the major contributing processes in darz and Nusse, 1998). Important to this study, the ma-defining this polarity is the subcellular trafficking of pro- jority of these conserved signaling pathway componentsteins to different membranes or compartments (re- are localized or enriched within the apical half of ex-viewed in Simons and Ikonen, 1997; Mostov et al., 2000). pressing cells. For example, the WG transmembraneFor transmembrane and secreted proteins, evidence ac- receptor, Frizzled (FZ), is enriched at the apical surfacecumulated to date suggests that the majority of this of imaginal disc cells (Park et al., 1994). The b-catenintrafficking occurs after protein synthesis via the directed homolog Armadillo (ARM), which translocates the signalmovement of specialized vesicles (Simons and Ikonen, to the nucleus, doubles as a structural component of1997; Ikonen and Simons, 1998; Yeaman et al., 1999). the apically localized adherens junctions (Orsulic and

Another process that may contribute significantly to Peifer, 1996). Disheveled (DSH) is enriched apically (Ya-protein trafficking and cell polarity is the subcellular nagawa et al., 1995; Torres and Nelson, 2000), and com-

ponents of the ARM-modifying complex, consisting oflocalization of transcripts prior to translation. Docu-E-Adenomatous polypopis coli (APC), Shaggy (SHG)/mented accounts of localized transcripts now numberZeste-White 3(ZW3), and Axin, are found at or near thewell over one hundred, and cover a large range of organ-apicolateral junctions (McCartney et al., 1999; Yu andBienz, 1999; Yu et al., 1999). WG itself is apically en-‡ To whom correspondence should be addressed (e-mail: h.krause@riched in and around WG-expressing cells (Gonzalez etutoronto.ca).al., 1991; Strigini and Cohen, 2000).§ Present address: The Hospital for Sick Children, Toronto, Ontario,

Canada. Here, we report that, as with many components of the

Cell198

Figure 1. Apical Localization of wg Tran-scripts Is Controlled by the 39 UTR

(A) and (C) show fluorescent in situ hybridi-zation detection of endogenous wg mRNAcaptured by confocal microscopy of stage8 embryos. (B) and (D) show correspondingexpression patterns of transcripts encodedby a wg-lacZ enhancer trap line. (A) and (B)are low magnification surface views and (C)and (D) are higher magnification opticalcross-sections through the ectodermal layer.Nuclei are green and transcripts red. Scalebars indicate 25 mm in (A) and (B) and 5 mmin (C) and (D).

wg signaling pathway, transcripts encoded by the wg fusion of the wg 39 UTR to lacZ resulted in apical tran-script localization (Figure 2A). Thus, the wg 39 UTR isgene are also localized apically. This localization is me-

diated by discrete elements within the wg 39 UTR. Redis- both necessary and sufficient for apical transcript local-ization.tribution of transcripts within the cell using heterologous

39 UTRs results in a redistribution of the protein and a To map the specific sequences responsible for apicallocalization within the 1098 nt wg 39 UTR, deletions weredramatic loss of signaling activity. The subcellular local-

ization of Wnt transcripts in other organisms suggests introduced into the lacZ-wg 39 UTR reporter, and thedeleted reporters tested for their ability to confer apicalthat this is a functionally conserved aspect of the WNT

signaling pathway. Indeed, this mechanism is likely to transcript localization in transgenic embryos (Figure 2B).These deletions defined two wg localization elementsplay a general role in the differential distribution of sig-

naling and transmembrane proteins in other polarized (WLEs), each of which is sufficient to confer apical tran-script localization. WLE1 is located between nucleotidesepithelia.60–178 and WLE2 is located between nucleotides 670–780 (Figure 2B). These elements may function differentlyResultsas localization conferred by WLE2 is more closely asso-ciated with the apical cortex than that conferred bywg Transcripts Are Localized Apically within CellsWLE1 (Figures 2C–2E). Differences in function are alsoof the Embryonic Ectodermsuggested by the lack of apparent similarity in sequencePrevious in situ hybridization studies with radioactiveor predicted secondary structure (not shown).probes suggested that wg transcripts were enriched

apically (Baker, 1987, 1988). Here, we use a recentlydeveloped fluorescent in situ hybridization technique Altering the Subcellular Distribution

of wg Transcripts(Hughes et al., 1996; Hughes and Krause, 1999) andconfocal microscopy to determine precisely the subcel- To examine the effect of transcript localization on WG

signaling, constructs expressing wg transcripts that lo-lular distribution of wg transcripts.Figure 1 shows the localization of wg transcripts in calize to different parts of the cell were made. The three

constructs made differ only in their 39 UTRs (Figuresthe ectoderm of a mid-stage 8 embryo (Figures 1A and1C; mRNA: red; nuclei: green). Transcripts encoded by 3A–3C). The first uses the endogenous wg 39 UTR, the

second a 39 UTR derived from the SV40 small t antigena lacZ reporter gene, expressed in the same cells undercontrol of the wg promoter, are shown for comparison gene, and the third a 39 UTR derived from the partner

of paired (ppa) gene (Raj et al., 2000). Each of the trans-(Figures 1B and 1D). Optical sections through individualstripes (Figures 1C and 1D) show that differences in the genes was placed under the control of a GAL4-depen-

dent promoter (Brand and Perrimon, 1993) and the vec-apparent width of each stripe are due to a difference insubcellular transcript localization. Transcripts encoded tors introduced into embryos to obtain transgenic flies.

Panels D–F in Figure 3 show the transcript distributionby the endogenous wg gene are confined to a smallarea just below the apical cell membrane (Figure 1C). patterns obtained when each of these transcripts is ex-

pressed under control of a wg-GAL4 driver in a wg mu-In contrast, lacZ transcripts expressed in the same cellsare distributed uniformly throughout the cytoplasm (Fig- tant background. As shown above, transcripts con-

taining the wg 39 UTR are localized apically (Figure 3D)ure 1D). We conclude that this localization is a transcript-specific and not a cell-specific property. Indeed, apical while transcripts containing the SV40 39 UTR are uni-

formly distributed (Figure 3E). In contrast, transcriptslocalization of wg transcripts is also observed in mostother polarized cells (localization in salivary gland cells containing the ppa 39 UTR are localized basally (Figure

3F). Note that the distribution of the ppa-tagged tran-is sometimes random; data not shown).Different portions of the wg transcript were tested script is not as tightly localized to the basal side of the

cell as full-length wg transcripts are to the apical surface.for their ability to confer apical localization in vivo toa nonlocalized lacZ transcript (Figure 2A). Transgenic Rather, the two distributions appear to be comple-

mentary.constructs with the wg 59 UTR and/or the wg ORF, fusedeither 59 (in-frame) or 39 to the lacZ sequence, yielded Prior to comparing the signaling activities of the pro-

teins made from these three transgenes, Western blotuniformly distributed transcripts (Figure 2A). However,

wg Transcript Localization199

Semi-quantitative RT-PCR analysis of the transgenictranscripts shows that transcription levels are alsoequivalent for each of the three lines in each of thematched sets (Figures 4E and 4F). Initial levels of proteinand RNA were also observed to be approximately equalwhen visualized in situ by immunocytochemistry and insitu hybridization (data not shown). We conclude thatthe 39 UTR swaps have little effect on the synthesis andstability of wg transcripts and protein. Posttranslationalmodifications also appear to be the same for each pro-tein, as each lane on the Western blot contains a similarset of bands equivalent in number, mobility, and relativeintensity (Figures 4A and 4B and data not shown).

The wg sequences in each of our constructs, bothcoding and noncoding, were sequenced prior to injec-tion. Each of the transgenes of the two matched setswas also recovered by PCR and their sequences recon-firmed. Therefore, any differences in protein activity thatmight be observed amongst matched lines must be at-tributed to differences in transcript localization.

WG Autoregulation Requires ApicalTranscript LocalizationWG facilitates its own expression via both autocrine andparacrine signaling pathways (Hooper, 1994; Manoukianet al., 1995; Yoffe et al., 1995). To test whether the local-ization of wg transcripts affects these activities, pulsesof wg construct expression were induced by crossingthe high set of UAS-wg transgenic flies to a heat shock-Gal4 line and subjecting 3- to 5-hr-old embryos to a 30min heat shock. Protein levels were assessed by West-ern blot analysis (Figures 5A and 5B).

Immediately following the heat pulse, each of theFigure 2. Two Regions within the wg 39 UTR Are Sufficient for Apical

matched transgenic lines produced the same amountLocalizationof protein. This was about three times the amount of(A) lacZ fusion constructs containing wg transcript 59, ORF, andendogenous WG expressed in heat-shocked controls. In39 regions are depicted. Transgenic constructs were expressed inthe apical transcript line, these levels rose about 5-foldembryos under the control of a ptc-GAL4 driver. The cytoplasmic

localization of each transcript in ectodermal cells is indicated on higher during the next half hour, and subsequently re-the right. mained at a high level. This increase in expression levels(B) Deletion constructs containing different portions of the wg 39 is due to the spatial expansion and intensification ofUTR are shown. Transgenic constructs were fused to the wg ORF

endogenous WG stripes (Noordermeer et al., 1992). Inand expressed under ptc-GAL4/UAS control. Solid black bars indi-the line with uniform wg transcript distribution, autoreg-cate regions remaining. Two minimal localization elements, termedulation also occurred but with slower kinetics. In con-WLE1 and WLE2, are indicated at the bottom. (C), (D), and (E) show

subcellular localization of the full-length (C), WLE1-containing (D), trast, the basal transcript line showed no further in-and WLE2-containing (E) transcripts depicted in (B). Transcripts are crease in WG expression levels 30 min after the heatred and cell outlines (anti-phosphotyrosine) are green. The intense pulse, and by 60 min, expression levels were similar tomedial spots in (C) and (E) are sites of nascent transcription in the

those seen in the heat shock control. Transcript levelsnucleus. (Bar 5 10 mm.)for each of the lines and each time point were alsomeasured using RT-PCR and NIH image, and confirmedanalysis (Figures 4A and 4B) was used to select trans-that, as with the ptc-GAL4-driven expression, each ofgenic lines that express equivalent levels of protein.the transgene mRNAs was expressed and turned overExpression of the transgenes was induced by crossingat equivalent rates (data not shown). Thus, we concludethe UAS-wg flies to ptc-GAL4 flies. These express GAL4that apical transcript localization is important for WGin the majority of ectodermal cells. Two matched setsautoregulation.of transgenic lines were selected, a “low”-expressing set

and a “high”-expressing set. Quantitation of the proteinWG Rescuing Activity Requires Apicallevels expressed by each of the fly lines in these setsTranscript Localization(Figures 4C and 4D) shows that, when the endogenousIn order to test the signaling activities of our differentiallyWG contribution is subtracted, the high lines expresslocalized transcripts in a more comprehensive fashion,about six times the levels of the low lines. Based on theeach of the constructs of our high- and low-expressingspatial differences between endogenous wg and ptc-matched sets was tested for their ability to rescue wg-GAL4-driven wg expression patterns, we estimate thatdependent segmental patterning. This was accom-the high lines express about half the levels of endoge-

nous WG protein on a per cell basis. plished by recombining the two sets of wg-expressing

Cell200

Figure 3. Differential Localization of wg Tran-scripts

wg constructs with different 39 UTRs weremade to localize transcripts to different re-gions of the cytoplasm. All three constructsexpress wg under the control of a GAL4-dependent promoter. The construct in (A)contains the full-length wg 39 UTR. The con-struct in (B) has the wg 39 UTR replaced bythe SV40 small t antigen gene 39 UTR. Theconstruct in (C) contains the 39 UTR of thepartner of paired gene in place of the wg 39

UTR. (D)–(F) show the subcellular distribu-tions of the transcripts encoded by eachtransgene when expressed in wg-expressingcells in a wg mutant background. Transcriptsare shown in red and cell outlines (a-spectrin oranti-phosphotyrosine) in green. (Bar 5 10 mm.)

transgenic lines into a wg null mutant background and tae are also found in cells nearby (Strigini and Cohen,1999; Pfeiffer et al., 2000; Figures 7A and 7B). In the wg-crossing these lines to flies that express GAL4 under

wg promoter control (Hays et al., 1997). As expected, expressing cells, these punctate bodies are thought torepresent both endocytic (Gonzalez et al., 1991) andthe apically localized transcript of the high apical line is

capable of restoring much of the naked cuticle that is exocytic vesicles (Pfeiffer et al., 2000).WG expressed from the apically localized transgenemissing in wg mutant embryos (Figure 6B). The incom-

plete nature of this rescue is most likely due to sub- transcript (Figures 7C and 7D) is distributed much thesame as in the wild-type control. However, protein ex-optimal levels of expression as compared to the endog-

enous wg gene (about 50%; Figures 4 and 7). In pressed from the uniformly distributed transcript (Fig-ures 7E and 7F) shows clear differences in distribution.comparison, the high uniform construct yields signifi-

cantly reduced rescuing activity (Figure 6C) and the high Although it still appears to be somewhat enriched in theapical cytoplasm of wg-expressing cells, there is lessbasolateral line very little rescuing activity (Figure 6D).

The low set of lines shows a similar trend, but with of the protein in these cells and more extending laterallyinto the middle of the segment. The difference in distri-substantially lower degrees of rescue (Figures 6E and

6G). Note that the low apical line yields more rescuing bution of protein translated from the basal transcript iseven more striking (Figures 7G and 7H). There is littleactivity than the high basal line despite expressing only

1/6th the levels of protein. Similar relative activities were detectable enrichment within the wg-expressing cells,and more of the protein extends laterally across theobserved using other GAL4 drivers and with other con-

structs that were not part of the matched sets (data not segment. Interestingly, this extracellular protein still ap-pears to be apically enriched. We conclude that proteinshown). We conclude that transcript localization within

apical cytoplasm is essential for robust signaling ac- synthesized basally is secreted more efficiently, diffusesmore rapidly within the extracellular matrix, or is lesstivity.effectively endocytosed.

WG Protein Distribution Is Affectedby Transcript Localization DiscussionIn order to help understand how transcript localizationaffects protein function, we looked to see if differences Subcellular Localization of wg Transcripts

Correlates with WG Signaling Activityin protein distribution, in and around wg-expressingcells, could be detected. Expression of the three trans- One of the paradigms of cell biology is that proteins

integrated within or secreted from a particular cell sur-genes in the high matched set was driven using a wg-GAL4 driver, and WG distributions were observed in face are generally synthesized from uniformly localized

transcripts. These transcripts bind to rough endoplas-a wg null background. The single-cell-wide wg stripes(Figures 3D–3F) serve as a point source from which mic reticulum near the nuclear perimeter, and sorting

begins after the protein-containing vesicles leave thediffusion of the protein, laterally and apically/basally,can be readily observed. Golgi stacks. Here, we show that the distribution of a

secreted protein can also be controlled at the level ofUsing confocal microscopy, most of the WG proteindetected in wild-type embryos is enriched in the apical transcript localization.

With the aid of high-resolution in situ hybridization,cytoplasm of wg-expressing cells (Figures 7A and 7B).Although the majority of this signal is diffuse, brightly we have shown that wg transcripts are enriched within

the apical cytoplasm of most epithelial cells examined.staining punctate bodies are also observed. Similar punc-

wg Transcript Localization201

Figure 4. Selection of Matched WG-Express-ing Transgenic Lines

(A) and (B) show representative Western blotsof WG protein expressed from matched api-cal, uniform, and basal wg-expressing trans-genic lines. (A) shows protein expressed froma low-expressing set of lines, and (B) showsprotein expressed from a high-expressing setof lines. Tubulin loading controls performedon the same blots are shown below. (C) and(D) are graphs showing phosphorimager-derived values for the relative levels of WGexpression in each of the transgenic lines ascompared to endogenous levels (normalizedto tubulin). Black bars are WG expression lev-els and hatched bars are tubulin expressionlevels. Error bars indicate standard devia-tions for four low set and ten high set blots.(E) and (F) show mRNA expression levels asdetected by RT-PCR. The agarose gel in (E)reflects mRNA levels expressed by the lowset of matched lines and the gel in (F) reflectsmRNA levels expressed by the high set. Tran-scripts encoded by the actin gene were alsoamplified as an internal control.

Indeed, the majority of this mRNA is concentrated just lation efficiency, cleavage, posttranslational modifica-tions, cofactor association, or trafficking of the newlybelow the apical cell surface, well away from the nuclear

envelope. When these transcripts are redistributed to synthesized protein.If localized translation of wg transcripts is affectingthe basal cytoplasm, the protein is efficiently translated

but shows little signaling activity. This loss in activity processing, these effects must be very subtle. Our West-ern blots revealed no obvious differences in translationcorrelates with changes in protein distribution. Less of

the protein is found within the apical cytoplasm of wg- efficiency, molecular weight, or glycosylation. Thus, api-cally and basally translated proteins are entering theexpressing cells and more is found in or around the cells

nearby. Preliminary expression studies at other stages secretory pathway with similar efficiencies. This rulesout the possibility of endoplasmic reticulum (ER) or golgiand in other tissues suggest that apical transcript local-

ization is a general requirement for wg signaling in epi- apparatus restriction to different parts of the cell. It doesnot, however, rule out the possibility of differential com-thelial cells.partmentalization along the apical/basal axis. For exam-ple, if Porcupine (PORC), an ER protein required for WGPotential Reasons for Transcript Localization

By controlling where in the cell transcripts are trans- exocytosis (van den Heuvel et al., 1993), were enrichedwithin an apical subregion of the ER, then protein syn-lated, localization could potentially affect a number of

protein properties. In the case of a secreted protein thesized basolaterally might be processed differentlyand/or enter a different secretory pathway together withsuch as WG, transcript localization might control the

organelle, or subregion thereof, where the protein would different cofactors. Either possibility could affect proteinsecretion, diffusion, or receptor binding. Recent supportultimately be synthesized. This, in turn, may affect trans-

Cell202

well as by members of the Hedgehog, TGFb, and FGFfamilies of signaling molecules (reviewed in Baeg andPerrimon, 2000; Christian, 2000; Selleck, 2000). Differen-tial distribution or modifications of these extracellularmatrix proteins along the apical/basal plane might ex-plain the observed differences in WG diffusion and activ-ity when translated basally. Although apical/basal differ-ences in the distribution of other extracellular matrixproteins are well documented (reviewed in Fessler andFessler, 1989; Gullberg et al., 1994; Murray et al., 1995),these particular proteins have yet to be analyzed.

Another possible consequence of basal WG secretionmay be the failure to associate effectively with FZ recep-tor complexes. As with many other receptor complexes,many of the proteins that bind and mediate the WGsignal appear to be closely associated and apically en-riched (Woods and Bryant, 1993; Yanagawa et al., 1995;Torres and Nelson, 2000; Orsulic and Peifer, 1996;McCartney et al., 1999; Yu and Bienz, 1999; Yu et al.,1999). WG may need to be secreted from the apicalsurface, perhaps even via a specific secretory pathway,in order to interact productively with these functionalprotein complexes. In imaginal discs, FZ itself is apicallylocalized (Park et al., 1994). FZ2, however, is not (Striginiand Cohen, 2000), indicating that the receptor itself neednot be the spatially limiting component.

In most wg-expressing tissues, layers of cells arefound basal to wg-expressing cells, and in some tissues,facing the apical surface as well. Although not ad-dressed here, apical localization may also provide amechanism to direct signals toward apically or laterallyopposed cells and away from basally opposed cells.This possible role as a determinant of signaling directionrequires further study.

Conservation of wg Transcript LocalizationTranscripts encoded by the Drosophila virilis wg gene

Figure 5. Autocrine wg Signaling Requires Apical Transcript Local- have been examined to see if apical transcript localiza-ization tion is conserved in this species. Despite an estimated(A) shows Western blots of WG expressed in heat shocked wild- evolutionary divergence of about 60 million years, andtype and HS-GAL4/UAS-wg transgenic constructs following a 30 the tendency of 39 UTRs to diverge rapidly in sequence,min heat shock. Protein was extracted from whole embryos at 0, this transcript is also localized apically (A. J. S. and30, and 60 min post heat shock. Actin loading controls are shown

H. M. K., unpublished data). Indeed, we find that the D.below each corresponding WG signal. (B) shows a plot of the relativevirilis wg 39 UTR is functional in D. melanogaster, andlevels of protein detected in the blots above. Levels are correctedelements with sequence similarity to the two wg localiza-for loading and are relative to WG expression levels in the heat

shocked control at 0 min. The signals shown were all within the tion elements, WLE1 and WLE2, exist in similar positionslinear range of phosphoimager detection. within the 39 UTR. Although conserved between species,

the sequences and predicted secondary structures ofWLE1 and WLE2 bear no resemblance to one another.Their sequences also fail to show significant homologyfor the possibility of different ER or golgi compartments

comes from the observation that yeast GPI-linked pro- to other sequences in the database, including thoseof other localized transcripts. Taken together with theteins are shuttled within a specific subset of vesicles

between these two organelles (Muniz et al., 2001). observation that wg transcripts colocalize to the sameparticles as other localized transcripts (Wilkie and Davis,Once WG is secreted, interactions with heparin sul-

fate-containing proteoglycans (HSPGs) within the extra- 2001 [this issue of Cell]), we surmise that transcript rec-ognition is mediated either by transcript-specific adap-cellular matrix are required for proper signaling (Bradley

and Brown, 1990; Jue et al., 1992; Reichsman et al., tors or by common adaptors that recognize similar sec-ondary structures.1996; Hacker et al., 1997; Lin and Perrimon, 1999; Tsuda

et al., 1999). These HSPGs are thought to function as This functional conservation of wg localization ele-ments in Drosophila further substantiates the impor-low affinity receptors that act at one of two levels, either

to facilitate binding to FZ receptors or to limit diffusion tance of wg transcript localization and suggests thatthis step in the pathway may be conserved in otheraway from FZ receptor complexes. These proteins are

also required for signaling by other WNT proteins as organisms. Indeed, a number of vertebrate wnt tran-

wg Transcript Localization203

Figure 6. Apical Transcript Localization IsRequired for wg Rescuing Activity

The UAS-regulated transgenes of our twomatched sets were expressed under controlof a wg-GAL4 driver in a wg mutant back-ground, and cuticles were prepared at theend of embryogenesis. (A) shows a wgcx4 cuti-cle in which the naked regions between denti-cle belts are deleted and the length of thecuticle is greatly reduced. (B) and (E) showcharacteristic rescue mediated by apically lo-calized wg transcripts, (C) and (F) by uni-formly localized transcripts and, (D) and (G)by basally localized transcripts. Cuticles onthe left (B, C, and D) are characteristic forrescue by the high-expressing set of lines andcuticles on the right (E, F, and G) for rescueby the low set of lines. The phenotypes shownare representative of the range produced byeach construct (n 5 100–200/construct).

scripts have been shown to be localized within oocytes. acting machinery required for localization (this study;Wilkie and Davis, 2001 [this issue of Cell]), this mode ofFor example, transcripts encoded by the Xwnt5 and

Xwnt11 genes of Xenopus are localized in the oocyte signaling molecule regulation may prove to be relativelycommon. This mode of protein trafficking may also bevegetal pole while those of X-Wnt8b are localized to the

animal pole (Ku and Melton, 1993; Moon et al., 1993; used for other types of polarity-conferring proteins suchas intercellular adhesion and cytoskeleton proteins.Cui et al., 1995). In Ascidians, the maternally expressed

HrWnt-5 transcript is localized to the posterior of earlyembryos (Sasakura et al., 1998). Recent advances in in Experimental Proceduressitu hybridization technologies should now permit the

Fly Stocks and P-Element Transformationsdetection of polarized transcript distributions in the so-In all cases, a w1118 stock was used as a wild-type control. The w;matic cells of these and other organisms.SbD2–3/TM6 Ubx, wgCX4, ptc-GAL4, wg-GAL4, and heat shock-GAL4(HS-GAL4) lines were obtained from the Bloomington DrosophilaStock Centre. All other stocks are described in Lindsley and ZimmTranscript Localization in Other(1992).Signaling Pathways

pUAST (Brand and Perrimon, 1993) derived constructs (see below)A number of signaling molecules in addition to WNTs were each independently sequenced and then introduced by micro-are known to be translated from localized transcripts. injection (Rubin and Spradling, 1982) into D2–3 embryos (Robertson

et al., 1988) to generate transgenic fly lines. Ten to twenty indepen-These include the Drosophila proteins Gurken, Sev-dent lines of each UAS-wg transgene were generated and all wereenless, Short gastrulation, Twisted gastrulation, and thetested for levels of protein and mRNA expression as describedXenopus protein Vg1 (Banerjee et al., 1987; Weeks andbelow. For the wg-GAL4 rescue experiments, UAS-wg transgenesMelton, 1987; Neuman-Silberberg and Schupbach, 1993;on the second chromosome were recombined with a second chro-

Francois et al., 1994; Mason et al., 1994). For some mosome containing a wgCX4 null allele.of these that are expressed in the oocyte, transcriptlocalization has been shown to be important (Neuman- Construction of wg/lacZ Fusion and wg Protein-Silberberg and Schupbach, 1994). The results presented Expressing Constructs

The wg ORF, 59 and 39 UTR sequences were tested for localizinghere show that transcript localization can also regulateactivity by subcloning each component downstream of the lacZsignaling protein distribution and activity in polarizedcoding sequence. The wg ORF was excised using a 59 NcoI siteepithelia, where it may serve as either an alternative orand a 39 AflII site located at the beginning of the 39 UTR. The 59a supplement to vesicle-mediated protein trafficking.UTR was subcloned using the 59 BamHI and 39 XhoI sites such that

Given the observations that at least 10% of transcripts the lacZ ORF would be in-frame with the wg signal peptide. Thehave the potential to be localized (Dubowy and Macdon- same fragment was also subcloned 39 to the lacZ sequence using

a 39 NcoI site. To subclone the wg 39 UTR, a 59 SacI site wasald, 1998) and that epithelial cells possess the trans-

Cell204

Figure 7. Effects of wg Transcript Localiza-tion on WG Protein Distribution

Each of the UAS transgenes of our high setwas expressed under control of a wg-GAL4driver in a wg mutant background. Panels onthe left show WG alone and panels on theright show composites of WG (red) and celloutlines (phosphotyrosine; green). (A) and (B)show endogenous WG localization and (C)–(H)the transgenic proteins. (C) and (D) show pro-tein expressed form apical transcripts, (E) and(F) from uniform transcripts, and (G) and (H)from basal transcripts. Each panel repre-sents a single 0.2 mm confocal section. (Bar 5

10 mm.)

introduced at the 59 end by PCR using the oligo ACGTGCTCACAGA Quantitation of wg Transgene Expression LevelsThe relative expression of WG within each of the transgenic linesTACTCGAGGGC (all oligos are written 59-39) and a 39 plasmid-spe-was examined by quantitative Western blotting using both actin andcific T7 primer. The 39 UTR was then excised with SacI and a BglIIb-tubulin signals as loading controls. For each line, 250 ptc-GAL4site located approximately 100 base pairs downstream from thevirgin females were crossed to 200 males of each UAS-wg line.predicted end of the cDNA (Nusse et al., 1984).These were allowed to mate for 3 days, and embryo collectionsTo facilitate fine mapping of the localization elements in the wgmade on the fourth and fifth days. For the heat shock time course39 UTR, a modular wg ORF fragment was first created from the wgexperiment, HS-GAL4 was used instead of ptc-GAL4. Embryos werecDNA. An XbaI site was introduced immediately upstream of thecollected for 2 hr, aged for 3 hr, and then heat shocked for 30 minstart codon and a BamHI site 39 to the stop codon using the primersin a 36.58C water bath. Total protein was extracted as describedGGTCTAGATGGATATCAGCT and GGGGATCCTTACAGACACGTGAbelow at 0, 30, and 60 min after heat shock. Embryos were firstand high-fidelity PCR using Pfu polymerase (Stratagene). Similarly,dechorionated in a 50% bleach solution, then flash-frozen in liquida modular 39 UTR fragment was made by introducing BamHI restric-nitrogen and transferred into 200 mL 13 SDS-PAGE loading buffertion sites at each end of the wg 39 UTR using the primers GGGGATCon ice. Embryos were then homogenized using a plastic pestle, andCACACTGCCCGCCT and GGGGATCCACTTGGCTTTTA. Large de-the resulting extract boiled for 5 min. Insoluble material was pelletedletions in the 39 UTR were made using internal AflII (position 109),using a 1 min microcentrifuge spin and 10 mL of the supernatantClaI (position 360), and HindIII (position 706) restriction sites. Smallerloaded onto a 10% SDS polyacrylamide gel. Resolved proteins weredeletions in the 1–360 construct were introduced using internal AluItransferred to 0.2 mm nitrocellulose membranes (S&S) using stan-(position 178) and BfrI (position 107) sites. The minimal localizationdard techniques and the membranes blocked overnight in PBTBelements were generated by PCR using the primers GGGATCCTG(1 3 PBS 1 0.2% Tween-20 and 0.2% skim milk powder). The wgTATGTGTTATGT and GGGGATCCAGAGCTTAAAGGGTT for WLE1,monoclonal antibody 4D4 (diluted 1:350; Brook et al., 1996), the

and GGGGATCCCAACAAATTCTTTTTT and GGGGATCCGCAACanti-b-tubulin polyclonal antibody E7 (diluted 1:600; Chu and Klym-

TAAAATCTT for WLE2.kowsky, 1989), and the anti-actin monoclonal antibody JLA20 (di-

The uniformly localized UAS-wg construct was created by substi- luted 1:400; Lin, 1981) were obtained from the Developmental Stud-tuting the wg 39 UTR with that of the SV40 small t antigen gene 39 ies Hybridoma Bank (developed under the auspices of the NICHDUTR, already present in the pUAST vector (Brand and Perrimon, and maintained by The University of Iowa, Department of Biological1993). The basolateral localizing UAS-wg construct was made by Sciences, Iowa City, Iowa). These were diluted in PBTB, incubatedfusing the wg ORF clone to a BamHI-HindIII fragment containing for 90 min at room temperature, and then washed six times for 5the 39 UTR of the partner of paired (ppa) gene (Raj et al., 2000). For min with PBTB. The membrane was then incubated in anti-mouseeach of the latter three constructs, the coding and 39 UTR sequences horseradish peroxidase conjugated secondary antibody (dilutedwere sequenced for errors introduced by cloning prior to injection 1:25,000; Pierce) for 90 min and washed six times in PBT (1 3 PBS 1into Drosophila embryos. These sequences were confirmed again 0.2% Tween-20) for a total of 30 min. Signals were developed usingafter isolation of the transgenic lines using the pUAST-specific prim- Supersignal-West-Dura chemiluminescent substrate (Pierce), anders CCGGAGTATAAATAGAGGCGCTTCG and GTCACACCACAGAA the WG-, tubulin-, and actin-specific bands quantified using aGTAAGGTTCC to amplify the transgenic sequences (Gloor et al., BioRad G360 Phosphoimager. Loadings were varied over a two1991). The amplified sequences were sequenced using the internal order of magnitude range to ensure that the signals detected and

quantified were in the linear range of the phosphoimager.primers GCTAAGCGAAAGCTAAGC and GATCCTCTAGAGGTAC.

wg Transcript Localization205

To measure transcript levels by RT-PCR, RNA was isolated from Baker, N.E. (1988). Transcription of the segment-polarity gene wing-less in the imaginal discs of Drosophila, and the phenotype of a4–6 hr ptc-GAL4/UAS-wg or HS-GAL4/UAS-wg embryos using an

RNeasy mini column (Qiagen), followed by a one-step RT protocol pupal-lethal wg mutation. Development 102, 489–497.(Qiagen), using the following primers: ATCATACCCCGTGTGTCAG Banerjee, U., Renfranz, P.J., Pollock, J.A., and Benzer, S. (1987).TGTGAGA (wg 59), ACCAGCAACCAAGTAAATCAACTGCA (UAS 59), Molecular characterization and expression of sevenless, a gene in-GGGCGTAATGTTGTTGGGTTCG (wg 39), AGCCAGCAGTCGTCTAA volved in neuronal pattern formation in the Drosophila eye. Cell 49,TCCAG (actin 59), and CAGCAACTTTCTTCGTCACACAT (actin 39). 281–291.During RNA extraction, the columns were incubated for 20 min with

Bashirullah, A., Cooperstock, R.L., and Lipshitz, H.D. (1998). RNADNase (Qiagen) before elution to prevent DNA contamination of the

localization in development. Annu. Rev. Biochem. 67, 335–394.RNA templates. RT-PCR was then performed on equal amounts of

Bocchinfuso, W.P., Hively, W.P., Couse, J.F., Varmus, H.E., andRNA, as described by the manufacturer. The RT reaction was al-Korach, K.S. (1999). A mouse mammary tumor virus-Wnt-1 trans-lowed to proceed at 508C for 30 min followed by 958C for 15 min.gene induces mammary gland hyperplasia and tumorigenesis inThe subsequent PCR cycling parameters were 40 cycles of: 948Cmice lacking estrogen receptor-alpha. Cancer Res. 59, 1869–1876.for 40 s, 558C for 40 s, and 728C for 60 s, followed by a final 10 min

at 728C. The resulting products were separated on 2% agarose gels, Bradley, R.S., and Brown, A.M. (1990). The proto-oncogene int-1photographed using a digital camera, and the relative intensities of encodes a secreted protein associated with the extracellular matrix.the bands quantified using NIH-Image for Windows 2000 (Scion). EMBO J. 9, 1569–1575.

Brand, A.H., and Perrimon, N. (1993). Targeted gene expression asImmunofluorescence and Fluorescent In Situ Hybridization a means of altering cell fates and generating dominant phenotypes.In situ hybridization, fluorescent detection of hybridized transcripts, Development 118, 401–415.and fluorescent detection of a-Spectrin in whole-mount embryos Broadus, J., Fuerstenberg, S., and Doe, C.Q. (1998). Staufen-depen-was performed as described previously (Hughes et al., 1996; Hughes dent localization of prospero mRNA contributes to neuroblastand Krause, 1999). The wg RNA probe used encompassed the entire daughter-cell fate. Nature 391, 792–795.ORF and was Digoxigenin labeled during run-off transcription as per

Brook, W.J., Diaz-Benjumea, F.J., and Cohen, S.M. (1996). Organiz-the manufacturer’s directions (Roche). Rabbit a-phosphotyrosine oring spatial pattern in limb development. Annu. Rev. Cell Dev. Biol.mouse anti-phosphotyrosine (NEB) antibodies were used at a dilu-12, 161–180.tion of 1:300 and 1:1000, respectively. Embryos were incubated atChristian, J.L. (2000). BMP, Wnt and Hedgehog signals: how far can48C overnight with either mouse anti-WG 4D4 1:10 or affinity-purifiedthey go? Curr. Opin. Cell Biol. 12, 244–249.rabbit anti-WG 1:50 diluted in PBTB. Embryos were then washed

four times at room temperature, 30 min each time, with PBTB. WG Chu, D.T., and Klymkowsky, M.W. (1989). The appearance of ace-antibody was detected using either an Alexa488 conjugated goat tylated alpha-tubulin during early development and cellular differen-anti-mouse or goat anti-rabbit secondary antibody (Molecular tiation in Xenopus. Dev. Biol. 136, 104–117.Probes), and the resulting signal was further amplified using an

Cui, Y., Brown, J.D., Moon, R.T., and Christian, J.L. (1995). Xwnt-Alexa488 conjugated donkey anti-goat antibody, each diluted

8b: a maternally expressed Xenopus Wnt gene with a potential role1:1000 in PBTB (Molecular Probes). Fluorescent antibodies were

in establishing the dorsoventral axis. Development 121, 2177–2186.added sequentially, each time incubating for 90 min and following

Dubowy, J., and Macdonald, P.M. (1998). Localization of mRNAswith four 20 min washes in PBTB. The anti-phosphotyrosine andto the oocyte is common in Drosophila ovaries. Mech. Dev. 70,anti-phosphotyrosine antibodies were detected using donkey anti-193–195.rabbit or donkey anti-mouse Cy5 secondary antibodies added 90Ephrussi, A., and Lehmann, R. (1992). Induction of germ cell forma-min before the final series of washes (Jackson Immunologicals).tion by oskar. Nature 358, 387–392.Nuclei were counterstained using propidium iodide (1 mg/ml; Sigma

Aldrich, Inc) added 5 min prior to the final washes. Images were Fessler, J.H., and Fessler, L.I. (1989). Drosophila extracellular matrix.obtained using a Leica TCS NT laser-scanning confocal microscope. Annu. Rev. Cell Biol. 5, 309–339.Each fluorochrome was scanned individually to avoid bleed-through

Francois, V., Solloway, M., O’Neill, J.W., Emery, J., and Bier, E.between channels. Images were subsequently combined using

(1994). Dorsal-ventral patterning of the Drosophila embryo dependsAdobe Photoshop 5.1.

on a putative negative growth factor encoded by the short gastrula-tion gene. Genes Dev. 8, 2602–2616.

Cuticle PreparationsGloor, G.B., Nassif, N.A., Johnson-Schlitz, D.M., Preston, C.R., andEmbryos were collected and cuticles prepared as previously de-Engels, W.R. (1991). Targeted gene replacement in Drosophila viascribed (Saulier-Le Drean et al., 1998). Approximately 400–500 cuti-P element-induced gap repair. Science 253, 1110–1117.cles were scored for each cross.Gonzalez, F., Swales, L., Bejsovec, A., Skaer, H., and Martinez Arias,A. (1991). Secretion and movement of wingless protein in the epider-Acknowledgmentsmis of the Drosophila embryo. Mech. Dev. 35, 43–54.

We wish to thank H. Lipshitz, C. Smibert, I. Davis, U. Tepass, and Gullberg, D., Fessler, L.I., and Fessler, J.H. (1994). Differentiation,various members of the Krause laboratory for comments on the extracellular matrix synthesis, and integrin assembly by Drosophilamanuscript. We also thank R. Nusse for wg cDNA, M. Weir for ppa embryo cells cultured on vitronectin and laminin substrates. Dev.cDNA, D. Branton for rabbit a-Spectrin, and S. Cohen and K. Cadigan Dyn. 199, 116–128.for a-WG antibodies. This work was supported by a grant from the Hacker, U., Lin, X., and Perrimon, N. (1997). The Drosophila sugarlessNational Cancer Institute of Canada. A. J. S. is supported by a K.M. gene modulates Wingless signaling and encodes an enzyme in-Hunter Fellowship in Cancer Research from the NCIC and G. dS. volved in polysaccharide biosynthesis. Development 124, 3565–by a Medical Research Council of Canada studentship. 3573.

Hays, R., Gibori, G.B., and Bejsovec, A. (1997). Wingless signalingReceived December 1, 2000; revised February 21, 2001.generates pattern through two distinct mechanisms. Development124, 3727–3736.

ReferencesHooper, J.E. (1994). Distinct pathways for autocrine and paracrineWingless signalling in Drosophila embryos. Nature 372, 461–464.Baeg, G., and Perrimon, N. (2000). Functional binding of secretedHughes, S.C., and Krause, H.M. (1999). Single and double FISHmolecules to heparan sulfate proteoglycans in Drosophila. Curr.protocols for Drosophila. Methods Mol. Biol. 122, 93–101.Opin. Cell Biol. 12, 575–580.

Baker, N. (1987). Molecular cloning of sequences from wingless, a Hughes, S.C., Saulier Le Dean, B., Livne-Bar, I., and Krause, H.M.(1996). Fluorescence in situ hybridization in whole-mount Drosophilasegment polarity gene in Drosophila: the spatial distribution of a

transcript in embryos. EMBO J. 6, 1765–1773. embryos. Biotechniques 20, 748–750.

Cell206

Ikonen, E., and Simons, K. (1998). Protein and lipid sorting from the progeny of wingless-expressing cells deliver the signal at a distancein Drosophila embryos. Curr. Biol. 10, 321–324.trans-Golgi network to the plasma membrane in polarized cells.

Semin. Cell Dev. Biol. 9, 503–509. Raj, L., Vivekanand, P., Das, T.K., Badam, E., Fernandes, M., Finley,Jue, S.F., Bradley, R.S., Rudnicki, J.A., Varmus, H.E., and Brown, R.L., Brent, R., Appel, L.F., Hanes, S.D., and Weir, M. (2000). Tar-A.M. (1992). The mouse Wnt-1 gene can act via a paracrine mecha- geted localized degradation of Paired protein in Drosophila develop-nism in transformation of mammary epithelial cells. Mol. Cell. Biol. ment. Curr. Biol. 10, 1265–1272.12, 321–328.

Reichsman, F., Smith, L., and Cumberledge, S. (1996). Glycosami-Kislauskis, E.H., Zhu, X., and Singer, R.H. (1997). beta-Actin messen- noglycans can modulate extracellular localization of the winglessger RNA localization and protein synthesis augment cell motility. J. protein and promote signal transduction. J. Cell Biol. 135, 819–827.Cell Biol. 136, 1263–1270.

Robertson, H.M., Preston, C.R., Phillis, R.W., Johnson-Schlitz, D.M.,Ku, M., and Melton, D.A. (1993). Xwnt-11: a maternally expressed Benz, W.K., and Engels, W.R. (1988). A stable genomic source ofXenopus wnt gene. Development 119, 1161–1173. P element transposase in Drosophila melanogaster. Genetics 118,Lawrence, J.B., and Singer, R.H. (1986). Intracellular localization of 461–470.messenger RNAs for cytoskeletal proteins. Cell 45, 407–415. Rubin, G.M., and Spradling, A.C. (1982). Genetic transformation ofLin, J.J.-C. (1981). Monoclonal antibodies against myofibrillar com- Drosophila with transposable element vectors. Science 218,ponents of rat skeletal muscle decorate the intermediate filaments 348–353.of cultured cells. Proc. Natl. Acad. Sci. USA 78, 2335–2339.

Sasakura, Y., Ogasawara, M., and Makabe, K.W. (1998). HrWnt-5:Lin, X., and Perrimon, N. (1999). Dally cooperates with Drosophila a maternally expressed ascidian Wnt gene with posterior localizationFrizzled 2 to transduce Wingless signalling. Nature 400, 281–284. in early embryos. Int. J. Dev. Biol. 42, 573–579.Lindsley, D.L., and Zimm, G.G. (1992). The Genome of Drosophila Saulier-Le Drean, B., Nasiadka, A., Dong, J., and Krause, H.M. (1998).melanogaster (San Diego, CA: Academic Press). Dynamic changes in the functions of Odd-skipped during early Dro-Long, R.M., Singer, R.H., Meng, X., Gonzalez, I., Nasmyth, K., and sophila embryogenesis. Development 125, 4851–4861.Jansen, R.P. (1997). Mating type switching in yeast controlled by

Selleck, S.B. (2000). Proteoglycans and pattern formation: sugarasymmetric localization of ASH1 mRNA. Science 277, 383–387.

biochemistry meets developmental genetics. Trends Genet. 16,Manoukian, A.S., Yoffe, K.B., Wilder, E.L., and Perrimon, N. (1995). 206–212.The porcupine gene is required for wingless autoregulation in Dro-

Simons, K., and Ikonen, E. (1997). Functional rafts in cell membranes.sophila. Development 121, 4037–4044.Nature 387, 569–572.

Mason, E.D., Konrad, K.D., Webb, C.D., and Marsh, J.L. (1994).Strigini, M., and Cohen, S.M. (1999). Formation of morphogen gradi-Dorsal midline fate in Drosophila embryos requires twisted gastrula-ents in the Drosophila wing. Semin. Cell Dev. Biol. 10, 335–344.tion, a gene encoding a secreted protein related to human connec-

tive tissue growth factor. Genes Dev. 8, 1489–1501. Strigini, M., and Cohen, S.M. (2000). Wingless gradient formation inthe Drosophila wing. Curr. Biol. 10, 293–300.McCartney, B.M., Dierick, H.A., Kirkpatrick, C., Moline, M.M., Baas,

A., Peifer, M., and Bejsovec, A. (1999). Drosophila APC2 is a cy- Takizawa, P.A., Sil, A., Swedlow, J.R., Herskowitz, I., and Vale, R.D.toskeletally-associated protein that regulates wingless signaling in (1997). Actin-dependent localization of an RNA encoding a cell-fatethe embryonic epidermis. J. Cell Biol. 146, 1303–1318. determinant in yeast. Nature 389, 90–93.Moon, R.T., Campbell, R.M., Christian, J.L., McGrew, L.L., Shih, J., Tomlinson, A., Bowtell, D.D., Hafen, E., and Rubin, G.M. (1987).and Fraser, S. (1993). Xwnt-5A: a maternal Wnt that affects morpho- Localization of the sevenless protein, a putative receptor for posi-genetic movements after overexpression in embryos of Xenopus tional information, in the eye imaginal disc of Drosophila. Cell 51,laevis. Development 119, 97–111. 143–150.Mostov, K.E., Verges, M., and Altschuler, Y. (2000). Membrane traffic Torres, M.A., and Nelson, W.J. (2000). Colocalization and redistribu-in polarized epithelial cells. Curr. Opin. Cell Biol. 12, 483–490. tion of Dishevelled and Actin during Wnt-induced mesenchymalMuniz, M., Morsomme, P., and Riezman, H. (2001). Protein sorting morphogenesis. J. Cell Biol. 149, 1433–1442.upon exit from the endoplasmic reticulum. Cell 104, 313–320.

Tsuda, M., Kamimura, K., Nakato, H., Archer, M., Staatz, W., Fox,Murray, M.A., Fessler, L.I., and Palka, J. (1995). Changing distribu- B., Humphrey, M., Olson, S., Futch, T., Kaluza, V., et al. (1999).tions of extracellular matrix components during early wing morpho- The cell-surface proteoglycan Dally regulates Wingless signallinggenesis in Drosophila. Dev. Biol. 168, 150–165. in Drosophila. Nature 400, 276–280.Neuman-Silberberg, F.S., and Schupbach, T. (1993). The Drosophila van den Heuvel, M., Harryman-Samos, C., Klingensmith, J., Perri-dorsoventral patterning gene gurken produces a dorsally localized mon, N., and Nusse, R. (1993). Mutations in the segment polarityRNA and encodes a TGF a-like protein. Cell 75, 165–174. genes wingless and porcupine impair secretion of the wingless pro-Neuman-Silberberg, F.S., and Schupbach, T. (1994). Dorsoventral tein. EMBO J. 12, 5293–5302.axis formation in Drosophila depends on the correct dosage of the

Weeks, D.L., and Melton, D.A. (1987). A maternal mRNA localizedgene gurken. Development 120, 2457–2463.

to the vegetal hemisphere in Xenopus eggs codes for a growthNoordermeer, J., Johnston, P., Rijsewijk, F., Nusse, R., and Law- factor related to TGF-b. Cell 51, 861–867.rence, P.A. (1992). The consequences of ubiquitous expression of

Wilkie, G., and Davis, I. (2001). Drosophila wingless and pair-rulethe wingless gene in the Drosophila embryo. Development 116,transcripts localize apically by dynein-mediated transport of RNA711–719.particles. Cell 105, this issue, 209–219.

Nusse, R., van Ooyen, A., Cox, D., Fung, Y.K., and Varmus, H. (1984).Wodarz, A., and Nusse, R. (1998). Mechanisms of Wnt signaling inMode of proviral activation of a putative mammary oncogene (int-1)development. Annu. Rev. Cell Dev. Biol. 14, 59–88.on mouse chromosome 15. Nature 307, 131–136.Woods, D.F., and Bryant, P.J. (1993). Apical junctions and cell signal-Orsulic, S., and Peifer, M. (1996). An in vivo structure-function studyling in epithelia. J. Cell Sci. Suppl. 17, 171–181.of Armadillo, the b-catenin homologue, reveals both separate and

overlapping regions of the protein required for cell adhesion and Yanagawa, S., van Leeuwen, F., Wodarz, A., Klingensmith, J., andfor Wingless signaling. J. Cell Biol. 134, 1283–1300. Nusse, R. (1995). The Dishevelled protein is modified by Wingless

signaling in Drosophila. Genes Dev. 9, 1087–1097.Park, W.J., Liu, J., and Adler, P.N. (1994). The frizzled gene of Dro-sophila encodes a membrane protein with an odd number of trans- Yeaman, C., Grindstaff, K.K., and Nelson, W.J. (1999). New perspec-membrane domains. Mech. Dev. 45, 127–137. tives on mechanisms involved in generating epithelial cell polarity.

Physiol. Rev. 79, 73–98.Pfeiffer, S., Alexandre, C., Calleja, M., and Vincent, J.P. (2000). The

wg Transcript Localization207

Yoffe, K.B., Manoukian, A.S., Wilder, E.L., Brand, A.H., and Perrimon,N. (1995). Evidence for engrailed-independent wingless autoregula-tion in Drosophila. Dev. Biol. 170, 636–650.

Yu, X., and Bienz, M. (1999). Ubiquitous expression of a Drosophilaadenomatous polyposis coli homolog and its localization in corticalactin caps. Mech. Dev. 84, 69–73.

Yu, X., Waltzer, L., and Bienz, M. (1999). A new Drosophila APChomologue associated with adhesive zones of epithelial cells. Nat.Cell Biol. 1, 144–151.